Ecological Indicators The effects

advertisement
Ecological Indicators 20 (2012) 143–150
Contents lists available at SciVerse ScienceDirect
Ecological Indicators
journal homepage: www.elsevier.com/locate/ecolind
The effects of drought and disturbance on the growth and developmental
instability of loblolly pine (Pinus taeda L.)
John H. Graham a,∗ , Jeffrey J. Duda b , Michelle L. Brown c , Stanley Kitchen d , John M. Emlen b ,
Jagadish Malol c , Elizabeth Bankstahl c , Anthony J. Krzysik e , Harold Balbach f , D. Carl Freeman c
a
Department of Biology, Berry College, Mount Berry, GA 30149, USA
U.S. Geological Survey, Western Fisheries Research Center, Seattle, WA 98115, USA
c
Department of Biological Sciences, Wayne State University, Detroit, MI 48202, USA
d
USDA Forest Service, Rocky Mountain Research Station, 735 North 500 East, Provo, UT 84606, USA
e
Prescott College, Prescott, AZ 86301, USA
f
U.S. Army ERDC-CERL, Champaign, IL 61826, USA
b
a r t i c l e
i n f o
Article history:
Received 22 August 2011
Received in revised form
15 December 2011
Accepted 4 March 2012
Key words:
Disturbance
Loblolly pine
Drought
Indicators
Developmental instability
a b s t r a c t
Ecological indicators provide early warning of adverse environmental change, helping land managers
adaptively manage their resources while minimizing costly remediation. In 1999 and 2000, we studied
two such indicators, growth and developmental instability, of loblolly pine (Pinus taeda L.) influenced
by mechanized infantry training at Fort Benning, Georgia. Disturbed areas were used for military training; tracked and wheeled vehicles damaged vegetation and soils. Highly disturbed sites had fewer trees,
diminished ground cover, warmer soils in the summer, and more compacted soils with a shallower Ahorizon. We hypothesized that disturbance would decrease the growth of needles, branches, and tree
rings, increase the complexity of tree rings, and increase the developmental instability of needles. Contrary to our expectations, however, disturbance enhanced growth in the first year of the study, possibly by
reducing competition. In the second year, a drought reduced growth of branches and needles, eliminating
the stimulatory effect of disturbance. Growth-ring widths increased with growing-season precipitation,
and decreased with growing-season temperature over the last 40 years. Disturbance had no effect on
tree-ring complexity, as measured by the Hurst exponent. Within-fascicle variation of current-year needle length, a measure of developmental instability, differed among the study populations, but appeared
unrelated to mechanical disturbance or drought.
© 2012 Elsevier Ltd. All rights reserved.
1. Introduction
Modification of the physical environment changes the growth
rate, density, age distribution, and genetic structure of populations,
and these changes cascade up the biological hierarchy to ecological
communities and ecosystems. Because many of these changes are
undesirable, the ability to detect change is useful. But what populations, and which aspects of those populations, should one examine?
Individual growth and homeostasis may be better indicators than
population size and density (Graham et al., 1993). Moreover, plant
species, especially woody plants, afford advantages over other taxonomic groups, such as animals and fungi (Freeman et al., 1993,
2003). Previously, we studied growth and developmental instability of herbs and shrubs in landscapes disturbed by military training
at Fort Benning in the Fall-line Sandhills of west-central Georgia,
∗ Corresponding author. Tel.: +1 706 290 2671; fax: +1 706 238 7855.
E-mail address: jgraham@berry.edu (J.H. Graham).
1470-160X/$ – see front matter © 2012 Elsevier Ltd. All rights reserved.
doi:10.1016/j.ecolind.2012.03.007
USA (Duda et al., 2003; Freeman et al., 2004a,b, 2005). Here, we
present the effects of military training and drought on growth and
developmental instability of a tree, loblolly pine (Pinus taeda L.). We
hypothesize that disturbance should decrease the growth of needles, branches, and tree rings, increase the complexity of tree-ring
widths, and increase the developmental instability of needles. We
found, however, that disturbance from military training actually
enhances growth of loblolly pine, while drought lessens it. Moreover, neither disturbance nor drought change tree-ring complexity
or increase developmental instability.
Trees are obvious and long-lived primary producers in forested
landscapes. Because of their long generations, they should adapt
slowly to disturbance (Smith and Donoghue, 2008, but see
Rosenheim and Tabashnik, 1991), retaining a history of growing conditions in their growth rings. Consequently, trees may be
better indicators of disturbance than herbs, particularly annuals,
which have shorter generations and adapt quickly to alterations in
their environment. Trees, on the other hand, are resilient, because
of their size and multiple storage compartments (King et al.,
144
J.H. Graham et al. / Ecological Indicators 20 (2012) 143–150
1999). Compartmentalization allows trees to buffer environmental
perturbations, generating time lags and consequent nonlinearity
(Zutter et al., 1998; Tang et al., 1999a,b). Moreover, trees are more
phenotypically plastic than herbs (Chambel et al., 2005).
By providing a cumulative, year-by-year history of growth, tree
rings reveal environmental stress decades to centuries in the past.
Moreover, one can correlate tree rings with past climate events
(Fritts, 1976; Carrer and Urbinati, 2006). Pest outbreaks, masting, and competition may also influence growth-ring width, but
the history of these events is normally unavailable. Thus, detecting the origins of stress amid the cacophony of noise is difficult.
Nevertheless, few taxa provide such information decades into
the past.
In contrast to tree rings, other indicators respond to stress on
finer time scales. Physiological indicators, such as water potential
and variable fluorescence, are useful for assessing environmental change and stress on scales of seconds to hours or weeks
(Sword et al., 1998; Tang et al., 1999a,b). Measures of developmental instability, such as fluctuating asymmetry, integrate the
response of trees over months to years, providing an intermediate time frame (Kozlov and Niemelä, 1999; Kozlov et al., 2002; see
Freeman et al., 2003, for a review) for a review. Together, growth
and developmental instability provide complementary indicators
of stress.
Because plastic responses, such as growth, are under selection
whenever stress appears, they should be sensitive to stress. Modification of leaf size, for example, alleviates the effects of drought
and shade (Raz et al., 2011). Thus, plasticity should be adaptive. In
addition, growth should be even more sensitive to stress; indeed
one can restate the popular definitions of stress directly in terms
of growth (Graham et al., 2010). For example, Grime (2001) defines
stress as anything that limits productivity (i.e., growth). Conversely,
Alekseeva et al. (1992) and Ozernyuk et al. (1992) define stress
as anything that dissipates energy (i.e., away from growth). See
Parsons (2005) for a recent review.
Developmental instability should be less sensitive to stress
than either growth or plasticity (Graham et al., 2010). Nevertheless, developmental instability has advantages; it is nonadaptive, whereas plasticity is adaptive (even decreased growth
can be adaptive if it increases fitness). Thus, developmental instability indicates when stress has adversely affected
development.
Developmental instability is estimated from measures of fluctuating asymmetry, the random deviations from perfect symmetry
(Graham et al., 2010). Most studies of fluctuating asymmetry deal
with bilaterally symmetrical traits, but deviations from other kinds
of symmetry can be used as well. Because plants are modular,
they are excellent subjects for fluctuating asymmetry. The variation among parts, such as the variation among pine needles within
a fascicle, is analogous to variation between right and left sides of
a bilaterally symmetrical trait.
Loblolly pine (Pinus taeda L.) is a shade-intolerant, early successional species. It produces three leaves (14–25 cm) per fascicle
and retains them for two to three years. Leaves are initiated in
the fall, elongate in the spring, and require four to five months to
mature (Tang et al., 1999a,b). We studied loblolly because it was
the dominant native pine at our sites.
We might expect a complicated response of loblolly to
disturbance—positive if disturbance removes competitors and negative if it damages the soil. Removal of understory herbs and shrubs
increases subsequent growth of loblolly (Haywood et al., 1997; Cain
and Shelton, 1998; Zutter et al., 1998; Tang et al., 1999a,b; Barnett
et al., 2002; Marino et al., 2002). Soil compaction, however, reduces
root growth, seedling survival, and shoot growth (Foil and Ralston,
1967) and overall performance is poor on shallow, eroded soils
(Fowells, 1965).
2. Materials and methods
2.1. Study sites
Fort Benning is a 73,533 ha military installation in the Fall-line
Sandhills near Columbus, Georgia (32◦ 20 N, 84◦ 41 W). Elevations
range from 61 to 225 m above sea level. Mean summer temperature
is 27 ◦ C, and mean winter temperature is 9 ◦ C; annual rainfall is
130 cm (Mason, 2003; Lozar, 2004). Upland forests at Fort Benning
are disturbed by mechanized infantry training (Dale et al., 2002;
Dilustro et al., 2002; Garten et al., 2003; Collins et al., 2006). The U.S.
Army uses highly disturbed areas for training with tracked vehicles.
Training here has removed much of the canopy, changed ground
cover, compacted the soils, and exposed bare soil. The infantry uses
other areas for training soldiers on foot, including navigation with
map and compass, and bivouacking. Finally, parts of the installation
are relatively pristine.
Fort Benning is managed for multiple uses, with large tracts of
land also used for timber management, recreation, and management of threatened and endangered species. From a management
perspective, the training throughout Fort Benning has created a
mosaic of habitats of differing quality. Nearly pristine habitat is
adjacent to heavily disturbed landscapes used for military training. This mosaic is suited for studies of landscape-level ecosystem
change.
We initially assigned several potential sites into three disturbance classes: highly (H), moderately (M), and lightly (L) disturbed
(Duda et al., 2003; Graham et al., 2004). Heavily disturbed areas
had reduced plant cover, large patches of bare soil, thinner soil Ahorizons, gullies, rock pedestals, and other signs of active erosion.
There was also abundant evidence of mechanized infantry training
(tank tracks, concertina wire, empty shell casings, etc.). Moderately
disturbed areas had an intact canopy, established ground cover, and
fewer signs of erosion. Nevertheless, gullies were present and there
was evidence of soil loss. There was also less evidence of vehicle use.
Lightly disturbed areas had minimal disturbance to soils, extensive ground cover, and the highest tree densities and canopy cover.
There were also no signs of recent erosion. Moreover, there was
little or no evidence of off-road vehicle and infantry use, except
on well-established trails. Within each of the three disturbance
classes, three sites were selected at random from upland mixedpine hardwoods forest in two adjacent third-order watersheds.
The sampling design was unbiased systematic-random, subject
to the qualifications regarding disturbance class. Such a design
insures adequate spatial coverage and representation, concurrent
with sample independence and an accurate estimation of P values. All sites were in Troup–Cowarts–Nankin Loamy-Sand soils
(NRCS, 2007). Two sets of sites were in the Bonham Creek watershed and one was in the Sally Branch watershed (both within HUC
031300030302).
Highly disturbed sites are used for mechanized infantry training with tracked and wheeled vehicles, foot solders, and bivouacs.
Moderately disturbed sites had previously experienced such use,
but at the time of the study had only foot traffic, with vehicles
restricted to roads and trails. Lightly disturbed sites had not experienced recent military training and are being managed for their
conservation value. All sites studied at Fort Benning experienced
timber harvest, agriculture, and grazing before the 1940s, when
these lands were added to the installation. The lightly disturbed
sites function in lieu of true controls. There were no completely
undisturbed sites of comparable character in the study area.
We selected research sites after a team of experienced field ecologists and resource managers had assessed damage to vegetation
and soils. A discriminant-function analysis of six habitat variables
effectively separated the three disturbance classes (Krzysik et al.,
2005). Lightly, moderately, and highly disturbed sites differed from
J.H. Graham et al. / Ecological Indicators 20 (2012) 143–150
each other in soil A-horizon depth and soil compaction. See Duda
et al. (2003), Graham et al. (2004, 2008), and Krzysik et al. (2005) for
descriptions of the nine sites, and photographs are available online
(Graham and Graham, 2008).
Rainfall data were obtained from the U.S. Army Corp
of Engineers Mobile District Water Management website
(http://water.sam.usace.army.mil/enhw.htm). The Fort Benning Natural Resources Branch provided data on the history
of prescribed burning and known wildfires. Temperature data
were obtained from the Southeast Regional Climate Center
(http://www.sercc.com/).
Based on available data from Fort Benning (1980–2002),
the average time between wildfires was 15.07 ± 8.19 years
(mean ± standard deviation), and the average time between prescribed burns was 3.84 ± 1.71 years (Graham et al., 2004). The
installation’s target for prescribed burns is a three year cycle.
Ignoring the source of the fire, the average time between burns
was 3.24 ± 1.66 years. There was no association between fire
(none, wildfire, prescribed) and disturbance class (2 = 7.162, df = 4,
P = 0.128).
2.2. Sampling
In July of 2001, we randomly sampled twenty loblolly at each
of the nine sites. For each tree, we recorded the circumference
breast height (CBH) and removed two branches 4–5 m above the
ground (where possible). From each branch, we measured the
annual growth (elongation) increment for 1999 and 2000 (±1 mm)
and collected three fascicles of needles from each annual internode
(1999 and 2000) per branch, selected at random from among all
fascicles produced on the branch in a given year. The three needles of each fascicle were separated and scanned with an Epson
Perfection 1200U Scanner at 125 dpi and Adobe Photoshop 5.0 software. Except for a small subsample, lengths were measured once
using CI-400 Computer Image Analysis Software (CID Incorporated,
Camas, Washington). Previous work with this system showed that
measurement error to both scanner and the CI-400 software is a
negligible (<1%) percent of the asymmetry variance (Freeman et al.,
2004a).
2.3. Growth
We collected a 0.51 cm diameter core from each loblolly, trying
to ensure that the core included pith. Cores were dried, mounted,
sanded along one side to reveal ring structure, and then scanned
(150 dpi) and measured using the same hardware and software as
above. Because incremental growth-ring width typically decreases
with plant age, we adjusted ring widths using a negative exponential of tree diameter (Fritts, 1976).
2.4. Developmental instability
Developmental instability was assessed as the within-fascicle
mean absolute deviation (MAD) of needle length. This measure of
variability is the expectation E of |xi − |, where the xi are the needle
lengths within a fascicle and is the average needle length for
that fascicle. E|xi − | is analogous to the most common index of
fluctuating asymmetry: E|l − r − |, where l is the value of a trait on
the left and r the value of a trait on the right. E|xi − | measures the
variation in needle length within a fascicle.
Measurement error is a consideration in studies of fluctuating
asymmetry (Graham et al., 2010). Because fluctuating asymmetry is a variance, or other measure of dispersion, measurement
error augments the true variation among needles in a fascicle.
The estimated measurement error (variance among replicates) was
0.002141, which was 32.3% of the asymmetry variance (variance
145
among needles within a fascicle = 0.004486). The signal-to-noise
ratio was moderate ( FA 2 / ME 2 = 2.095).
2.5. Statistical analysis
We used repeated-measures analysis of variance (ANOVA) to
test the hypotheses that annual growth increment, needle length,
and needle asymmetry differed among disturbance classes. Disturbance class was a fixed effect and year was a repeated effect. Sites
were random effects nested within disturbance classes, and individual trees were nested within sites. For annual growth increment,
branches were nested within individual trees. For needle length and
asymmetry, fascicles were nested within branches within individual trees.
We used analysis of covariance (ANCOVA) to test the hypothesis
that adjusted tree-ring width differed among disturbance classes.
Average growing-season rainfall and temperature were covariates.
We used nested ANOVA to test the hypotheses that age and circumference (CBH) differed among disturbance classes. Disturbance
class was a fixed effect, with sites nested within disturbance class
and trees nested within sites.
The dependent variables required transformations to meet the
assumptions of the ANOVA. Because age generally fits a Poisson distribution (i.e., counts of randomly occurring objects), we used the
square root of age in all analyses. Because growth is a multiplicative
process, we used log10 (CBH + 1) as the dependent variable for circumference. For needle length, we used a Box–Cox transformation
([x − 1]/), where = 0.7. This was especially important for needle
asymmetry to prevent either positive or negative size scaling (see
Graham et al., 2010). The raw needle lengths exhibited positive
size scaling (regression of var(xi ) on E(xi ), F = 6.08, df = 1, P < 0.02)
while the coefficient of variation exhibited negative size scaling
(F = 141.40, df = 1, P < 0.001). A Box–Cox transformation with a value
of = 0.7 did not exhibit size scaling (F = 0.010, df = 1, P > 0.90).
Whenever there were more than two means, we used
Tamhane’s T2 post hoc test (Tamhane, 1979). Tamhane’s T2 is a
conservative test that does not assume equality of variances.
To further explore the effect of disturbance, we used growthring data from the four oldest trees (35 to over 50 years of age)
at each site to examine the pattern of growth from 1948 to 2001
by computing the Hurst exponent (H) (Beran, 1994). The range of
observations R in a time series depends upon the time period . The
relationship fits the Hurst equation
R
= (a)H ,
S
where R is the range, S is the standard deviation of the range, a
is a constant, is the time period, and H is the Hurst exponent.
The rescaled range, R/S, is a dimensionless ratio. The Hurst exponent is a measure of the rate at which R/S increases with . The
Hurst exponent is related to the fractal dimension of the time series:
D = 2 − H.
The Hurst exponent differs depending upon the process producing the growth. Random processes have no memory and generate a
random walk (e.g., Brownian motion). Random walks have a Hurst
exponent of 0.5. Processes that demonstrate inertia, or persistence,
tend to continue moving in the direction they are moving rather
than shifting back. Such persistent processes have a Hurst exponent of greater than 0.5. In contrast, processes exhibiting a higher
likelihood of changing direction have Hurst exponents less than
0.5. With respect to tree-ring growth, given no internal buffering (i.e., autocorrelation), the Hurst exponent should be 0.5 if
favorable years occur at random. Prolonged periods of good (or
bad) growing seasons should yield an exponent greater than 0.5,
whereas if good years are followed by bad years, the Hurst exponent should be less than 0.5. To the extent that individuals buffer
146
J.H. Graham et al. / Ecological Indicators 20 (2012) 143–150
Fig. 1. Rainfall (in cm) for the water year (March–October) of the study period. The
dashed line is the 55-year average.
themselves from environmental change, fluctuations in tree-ring
width should be dampened relative to fluctuations in environmental conditions. This will manifest in a tree-ring Hurst exponent
exceeding those of the climate conditions upon which growth
depends (temperature and rainfall). A priori, we expected that
plants at the lowest disturbance would display the highest Hurst
exponents, since these plants should best buffer environmental
fluctuations, whereas those at the heavily disturbed sites should
display less persistent growth (Emlen et al., 1993; Beran, 1994).
3. Results
3.1. Rainfall
We used the record of rainfall at Fort Benning (Fig. 1) to evaluate conditions for growth and developmental instability of loblolly
pine in the three years from 1998 to 2000. Over the 55-year
period, average rainfall for the water year (October–April) was
72.9 cm. For the period 1948–2001, rainfall had a Hurst exponent
(H) of 0.303 ± 0.0105 (mean ± 95% confidence interval, r2 = 0.850,
P < 0.001).
3.2. Temperature
Average annual temperatures, especially those during the growing season (March–October), also have the potential to influence
growth and developmental instability of loblolly pine (Fig. 2). The
Fig. 3. Age and circumference breast height (CBH) by site within disturbance class.
(A) Square root of age, and (B) log10 (CBH + 1). Error bars are 95% confidence intervals.
L1, L2, and L3 are light disturbance. M1, M2, and M3 are moderate disturbance. H1,
H2, and H3 are heavy disturbance.
lowest average temperature during the growing season occurred
in 1961 (21.2 ◦ C), whereas the highest average temperature during the growing season occurred in 2002 (23.6 ◦ C), the last year of
record during the study. The average temperature, over all growing
seasons, was 22.3 ◦ C (standard deviation = 1.0).
The average annual temperature was correlated with the
growing season annual temperature (r = 0.624, df = 50, P < 0.0001),
and uncorrelated with precipitation during the growing season
(r = 0.035, df = 50, P > 0.75). The average annual temperature had a
Hurst exponent of 0.752 ± 0.005 (mean ± 95% confidence interval,
r2 = 0.992, P < 0.001). The average annual temperature during the
growing season had a Hurst exponent of 0.75 ± 0.005 (mean ± 95%
confidence interval, r2 = 0.992, P < 0.001).
3.3. Loblolly pine age and size
Neither age (square root of age in years) nor circumference
(log10 [CBH + 1]) differed among the three disturbance classes (age:
F2, 6 = 0.295, P > 0.75; CBH: F2, 6 = 2.38, P > 0.17). Much of the variation in age and CBH was among sites within disturbance classes
(age: F6, 156 = 3.42, P < 0.005; CBH: F6, 170 = 3.25, P < 0.01). Ignoring
disturbance class, there were differences in age (F8, 156 = 2.743,
P < 0.01) and CBH (F8, 170 = 4.335, P < 0.001) among the nine sites
(Fig. 3). Trees at M2 were older than those at M1, M3, and H1
(Tamhane T2, all P < 0.05). Trees at L1 had smaller CBH than those
at either H3 (Tamhane T2, P < 0.005) or M3 (Tamhane T2, P < 0.05).
Trees at H1 were marginally smaller than those at H3 (Tamhane T2,
P = 0.05).
Circumference, standardized by age, is a measure of long-term
growth. When age was used as a covariate in an analysis of
covariance (ANCOVA), there was still no difference in CBH among
disturbance classes (F2, 6 = 2.275, P > 0.180).
3.4. Branch annual growth increment
Fig. 2. Average growing season (March–October) temperature in ◦ C. The dashed line
is the 55-year average.
The annual growth increment (log10 [internode length + 1])
was greater in 1999 than in 2000 (repeated measures ANOVA,
F1, 169 = 48.75, P < 0.001) (Fig. 4). Moreover, there were differences in
growth increment among disturbance classes (repeated measures
J.H. Graham et al. / Ecological Indicators 20 (2012) 143–150
Fig. 4. Internode length by year and disturbance class. Error bars are 95% confidence
intervals.
ANOVA, F2, 6 = 13.878, P < 0.001). Trees at the highly disturbed sites
had greater annual growth than trees at the lightly disturbed sites
(Tamhane T2, P < 0.025) and they had marginally greater growth
than trees at the moderately disturbed sites (Tamhane T2, P < 0.06).
There was a disturbance-by-year interaction as well (repeated measures ANOVA, F2, 169 = 7.54, P < 0.001). The differences in growth
among disturbance classes were pronounced in 1999, but not in
2000 (Fig. 4).
Variation among sites within disturbance class was not nil
(repeated measures ANOVA, F6, 165 = 28.58, P < 0.001). Indeed, it
accounts for much of the variation in yearly branch growth. Moreover, trees within sites within disturbance class differed as well
(repeated measures ANOVA, F165, 169 = 2.42, P < 0.001).
3.5. Average needle length
Average needle length (Box–Cox transformation, = 0.7) differed between years (repeated measures ANOVA, F1, 170 = 505.12,
P < 0.001). Needle length in 2000 was only 80.2% of what it was
in 1999. It also increased with increasing disturbance (repeated
measures ANOVA, F2, 6 = 4.021, P < 0.025, Fig. 5), but the post
hoc comparisons among disturbance classes were insignificant
(Tamhane’s T2, P > 0.30). There was also variation in needle
length among sites within disturbance classes (repeated measures
147
Fig. 6. Needle asymmetry [mean absolute deviation, with Box–Cox transformation ( = 0.7) of needle lengths] by site and disturbance class. Error bars are 95%
confidence intervals.
ANOVA, F6, 165 = 4.957, P < 0.001) and among plants within sites
(repeated measures ANOVA, F165, 170 = 3.428, P < 0.001).
3.6. Developmental instability of needles
The within-fascicle mean absolute deviation (MAD) of needle length did not vary between years (repeated measures
ANOVA, F1, 170 = 0.198, P > 0.65) and disturbance classes (repeated
measures ANOVA, F2, 170 = 0.463, P > 0.60). Moreover, the interaction of years and disturbance was insignificant (repeated
measures ANOVA, F2, 170 = 0.365, P > 0.33). There was, however,
variation among sites within disturbance classes (repeated measures ANOVA, F6, 165 = 3.261, P < 0.01), but not among plants within
sites (repeated measures ANOVA, F165, 170 = 1.214, P > 0.10). In post
hoc comparisons of the nine sites (repeated measures ANOVA,
F8, 165 = 2.58, P < 0.015), three comparisons were marginally significant (Tamhane T2, P < 0.09). Sites H1, M3, and L2 all showed greater
variation in within-fascicle needle length than M2 (Fig. 6).
3.7. Growth ring width
The logarithm of the adjusted ring growth declined as the average growing-season temperature increased (F1, 4176 = 4.67, P < 0.05)
and as growing-season precipitation decreased (F1, 4176 = 3.92,
P < 0.05). After removing the effects of temperature and precipitation, ring width increased as disturbance increased (F2, 6.072 = 5.52,
P < 0.05, Fig. 7).
We examined the adjusted width of the growth rings over time
to determine if disturbance influenced the persistence of growth
patterns, as measured by the Hurst exponent (Emlen et al., 1993).
As expected, the tree-ring Hurst exponent exceeded that for both
rainfall and temperature. More importantly, it did not differ across
the disturbance gradient (Fig. 7).
4. Discussion
Fig. 5. Box–Cox transformation ( = 0.7) of needle length by year and disturbance
class. Error bars are 95% confidence intervals.
We evaluated the effects of military training on growth and
developmental instability of loblolly pine at Fort Benning, Georgia.
Our data span several temporal scales. The annual increment of
branch growth, needle length, and the developmental instability of
needle length provide scales of less than two years. Tree growth
rings and the Hurst exponent allow us to look at scales of fifteen to
over fifty years. Overall, soil disturbance had little negative effect
on loblolly performance. It enhanced growth, but had no effect
on within-fascicle needle-length variation. Drought, on the other
148
J.H. Graham et al. / Ecological Indicators 20 (2012) 143–150
Loblolly also does better on the edges of forest fragments.
McDonald and Urban (2004) found that individuals growing on the
edges grew faster than those in the interior. Nevertheless, the effect
was small. Forest edges explained less of the variance in growth rate
than did soil texture, soil nutrients, and topography.
Soil ammonia concentrations were highest at the undisturbed
sites and lowest at the highly disturbed sites, while soil nitrate
followed an opposite pattern (D.A. Kovacic, University of Illinois,
unpublished data). Taken together, these suggest that competition from understory vegetation, along with soil organic matter,
is mediating the nutrients available for loblolly growth. Because
the soil carbon compartment is lacking at the highly disturbed
sites (DeBusk et al., 2005; Maloney et al., 2008), its ability to
influence nutrients is likely reduced, resulting in a simplified
two-compartment system. Such simplification should lead to pronounced oscillations and reduced complexity in growth over time,
though we were unable to detect such reductions in complexity.
Pinyon pines having pronounced oscillations have increased mortality (Ogle et al., 2000).
The interaction of year and disturbance on annual increment of
branch growth suggests that highly disturbed sites are less resilient
to drought than lightly and moderately disturbed sites. In 1999, the
trees had been exposed to one year of drought. By 2000, they had
been exposed to two years of drought.
4.3. The effects on buffering of stress
Fig. 7. Adjusted tree-ring widths and Hurst exponent by disturbance class. Error
bars are 95% confidence intervals.
The Hurst exponents for growth rings in loblolly (H = 0.80–0.82)
exceeded those for temperature (H = 0.75) and precipitation
(H = 0.30), indicating that loblolly’s growth is more persistent than
are the patterns of these climatic variables. Trees do indeed buffer
themselves against the vicissitudes of nature.
4.4. The effect of year
hand, had a negative effect on loblolly growth and tree-ring width,
though it did not influence needle-length variation.
4.1. Rainfall and temperature
Hurst exponents for the time series of rainfall and temperature at Fort Benning indicate long-term climatic patterns. Rainfall
was anti-persistent (H < 0.5), frequently swinging between moist
and dry years. Especially relevant to our study, a drought began
in 1999 and continued through 2000. Temperature was persistent (H > 0.5), moving in the same, warmer, direction rather than
swinging between temperature extremes.
We have data on needle length and branch length for two years:
1999 and 2000. A drought began in 1999 and intensified in 2000.
Needle length in 2000 was 80.2% of what it was in 1999 at all disturbance levels; so there was a clear effect of year on needle growth.
Similarly, branch growth was less in 2000 than in 1999. The interaction of year and branch growth suggests that drought blunts the
positive effects of disturbance on growth. McDonald and Urban
(2004) found a similar response in that the difference in growth
between loblolly on a forest edge and those in the forest interior
was greater on drier soils. Houle and Delwaide (1991) found that
loblolly growing on rock outcrops were more sensitive to drought
than those on nearby Piedmont.
4.5. Developmental instability
4.2. Growth
The annual increment of branch growth, needle length, and the
adjusted width of growth rings all increased with military disturbance. Thus, disturbance positively influenced plant growth.
Consequently, disturbance, per se, is not a stress for loblolly, though
frequent disturbance does reduce the ability of the pines to grow
during prolonged droughts (Zahner and Saucier, 1989; Chen et al.,
1994; Oberhuber et al., 1998). The most likely explanation for this
surprising outcome is that disturbance at Fort Benning removes
understory competitors, thereby reducing stress.
There is evidence that understory plants can depress growth of
loblolly through competition (Haywood et al., 1997; Marino et al.,
2002; Martin and Jokela, 2004). Heavily disturbed sites at Fort
Benning have more bare ground than the lightly and moderately
disturbed sites (Duda et al., 2003; Krzysik et al., 2005). They also
have less woody ground cover and less canopy cover.
We used the within-fascicle mean absolute deviation (MAD) of
needle lengths to assess developmental instability. Neither year nor
disturbance class had an effect on developmental instability of needle length. There were differences in MAD among the nine sites,
with H1, M3, and L2 having more variable needle lengths within
a fascicle than those at M2. We are unsure what is responsible for
these differences.
4.6. General comments
Our highly disturbed sites were obviously degraded. The bare
ground and erosion suffice to convince. But the pattern of loblolly
growth was unexpected: internode length, needle length, and treering width all improved as disturbance increased. Clearly, growth
does not provide a useful diagnostic indicator of stress for natural
resource managers responsible for Fort Benning ecosystems.
J.H. Graham et al. / Ecological Indicators 20 (2012) 143–150
Elsewhere, we reported the developmental instability of three
other species sampled at the same sites, two herbs, finger rot
(Cnidoscolus stimulosus [Michx.] Engelm. & Gray) and morning glory
(Ipomoea pandurata [L.] G.F.W. Mey), and a shrub, sumac (Rhus
copallinum L.) (Duda et al., 2003; Freeman et al., 2004a,b, 2005). This
allows us to compare developmental instability across species. Finger rot had more asymmetric leaves at the highly disturbed sites.
Morning glory, in contrast, had the lowest leaf asymmetry in the
moderately disturbed sites.
We also examined net photosynthesis, transpiration, variable
fluorescence, and water stress of sumac at these same sites (Duda
et al., 2003; Freeman et al., 2004b). These are our short-term indicators of stress. Like growth, net photosynthesis of sumac and
morning glory increased with increasing disturbance, especially
after fire. There was no clear disturbance effect on water stress
or variable fluorescence. The fluorescence did indicate that all sites
were stressful; however, there was no pattern to the amount of
stress (Duda et al., 2003; Freeman et al., 2004b).
Of necessity, we studied species found at all nine sites.
Unfortunately, the landscape is a mosaic of disturbance, and individuals may occur only in favorable microhabitats surrounded
by degraded habitat, precluding any real measure of stress
(Duda et al., 2003).
Another explanation for the lack of increase in developmental
instability with disturbance of loblolly could be that the ecosystem has adjusted to military activities, such that the disturbance
no longer constitutes a stress. This simple conclusion is suggested
by the increased growth in the disturbed areas, but is complicated
by the higher developmental instability with high disturbance in
both finger rot and sumac, and by the apparent loss of buffering
capacity in loblolly with increased disturbance, as indicated by the
reduction in growth in 2000 relative to 1999 at the most disturbed
sites.
5. Conclusions
None of our indices of growth indicated that disturbance by
military training was stressful to loblolly pine. Contrary to expectations, disturbance improved branch growth, needle length, and
ring width, possibly by alleviating competition. Drought, on the
other hand, depressed growth, thereby indicating these events to be
stressful. These results are consistent with the idea that indicators
of stress are useful only while stress-induced ecosystem changes
are actually taking place. Once a change has occurred, the system
adjusts to the new conditions, which cease being stressful. Unpredictable changes, however, such as drought, permit only a lagging
response, and so continue to be stressful.
Acknowledgments
We thank Hugh Westbury (U.S. Army ERDC-CERL) for coordinating our fieldwork at Fort Benning, and Pete Swiderek, Natural
Resources Branch Chief, for his assistance in site selection, and John
Brent, Environmental Management Division Chief, for research
support. We also thank Lisa Ceravolo, Cathy Chamberlin-Graham,
Michelle Hobig, Jennifer Murray, and Michelle Nilson for assisting
with data collection. This research was funded by SERDP (Strategic Environmental Research and Development Program) Project
CS-1114B, Development of Ecological Indicator Guilds for Land
Management. SERDP is a joint cooperative research program of the
Department of Defense, Department of Energy, and U.S. Environmental Protection Agency. These studies comply with the current
laws of the United States of America. Use of trade names is for convenience of the reader and does not constitute endorsement by the
U.S. government.
149
References
Alekseeva, T.A., Zinichev, V.V., Zotin, A.I., 1992. Energy criteria of reliability and
stability of development. Acta Zool. Fenn. 191, 159–165.
Barnett, M.M., Rideout, S., Oswald, B.P., Farrish, K.W., Williams, H.M., 2002. Growth
response from herbicide, prescribed fire, and fertilizer treatments in midrotational loblolly pine: first-year response. General Technical Report SRS 48. U.S.
Department of Agriculture, Forest Service, Southern Research Station, Asheville,
NC, pp. 143–146.
Beran, J., 1994. Statistics for Long-Memory Processes. Chapman & Hall, Boca Raton,
FL, 315 pp.
Cain, M.D., Shelton, M.G., 1998. Viability of litter-stored Pinus taeda L. seeds after
simulated prescribed winter burns. New Forest 16, 1–10.
Carrer, M., Urbinati, C., 2006. Long-term change in the sensitivity of tree-ring growth
to climate forcing in Larix decidua. New Phytol. 170, 861–872.
Chambel, M.R., Climent, J., Alía, R., Valladares, F., 2005. Phenotypic plasticity: a useful
framework for understanding adaptation in forest species. Investigación agraria:
Sistemas y recursos forestales 14, 334–344.
Chen, C.W., Tsai, W.T., Gomez, L.E., 1994. Modeling responses of ponderosa pine to
interacting stresses of ozone and drought. Forest Sci. 40, 267–288.
Collins, B., Sharitz, R., Madden, K., Dilustro, J., 2006. Comparison of sandhills and
mixed pine-hardwood communities at Fort Benning, Georgia. Southeast Nat. 5,
93–102.
Dale, V.H., Beyeler, S.C., Jackson, B., 2002. Understory vegetation indicators of anthropogenic disturbance in longleaf pine forests at Fort Benning, Georgia, USA. Ecol.
Ind. 1, 155–170.
DeBusk, W.F., Skulnick, B.L., Prenger, J.P., Reddy, K.R., 2005. Response of soil organic
carbon dynamics to disturbance from military training. J. Soil Water Conserv.
60, 163–171.
Dilustro, J.J., Collins, B.S., Duncan, L.K., Sharitz, R.R., 2002. Soil texture, land-use intensity, and vegetation of Fort Benning upland forest sites. J. Torrey Bot. Soc. 129,
289–297.
Duda, J.J., Freeman, D.C., Brown, M.L., Graham, J.H., Krzysik, A.J., Emlen, J.M., Zak, J.C.,
Kovacic, D.A., 2003. Estimating disturbance effects from military training using
developmental instability and physiological measures of plant stress. Ecol. Ind.
3, 251–262.
Emlen, J.M., Freeman, D.C., Graham, J.H., 1993. Nonlinear growth dynamics and the
origin of fluctuating asymmetry. Genetica 89, 77–96.
Foil, R.R., Ralston, C.W., 1967. The establishment and growth of loblolly pine
seedlings on compacted soils. Soil Sci. Soc. Am. J. 31, 565–568.
Fowells, H.A., 1965. Silvics of forest trees of the United States. U.S. Department of
Agriculture, Forest Service, Agriculture Handbook No. 271, 762 pp.
Freeman, D.C., Brown, M.L., Duda, J.J., Graham, J.H., Emlen, J.M., Krzysik, A.J., Balbach, H., Kovacic, D.A., Zak, J.C., 2004a. Developmental instability in Rhus
copallinum L.: multiple stressors, years, and responses. Int. J. Plant Sci. 165,
53–63.
Freeman, D.C., Brown, M.L., Duda, J.J., Graham, J.H., Emlen, J.M., Krzysik, A.J., Balbach,
H., Kovacic, D.A., Zak, J.C., 2004b. Photosynthesis and fluctuating asymmetry
as indicators of plant response to soil disturbance in the Fall Line Sandhills of
Georgia: a case study using Rhus copallinum and Ipomoea pandurata. Int. J. Plant
Sci. 165, 805–816.
Freeman, D.C., Brown, M.L., Duda, J.J., Graham, J.H., Emlen, J.M., Krzysik, A.J., Balbach,
H., Kovacic, D.A., Zak, J.C., 2005. Leaf fluctuating asymmetry, soil disturbance and
plant stress: a multiple year comparison using two herbs, Ipomoea pandurata and
Cnidoscolus stimulosus. Ecol. Ind. 5, 85–95.
Freeman, D.C., Graham, J.H., Emlen, J.M., 1993. Developmental stability in plants:
symmetries, stress and epigenesis. Genetica 89, 97–119.
Freeman, D.C., Graham, J.H., Emlen, J.M., Tracy, M., Hough, R.A., Alados, C.L., Escós,
J., 2003. Plant developmental instability: new measures, applications, and regulation. In: Polak, M. (Ed.), Developmental Instability: Causes and Consequences.
Oxford University Press, New York, pp. 367–386.
Fritts, H.C., 1976. Tree Rings and Climate. Academic Press, New York, 567 pp.
Garten, C.T., Ashwood, T.L., Dale, V.H., 2003. Effect of military training on indicators
of soil quality at Fort Benning, Georgia. Ecol. Ind. 3, 171–179.
Graham, C.C., Graham, J.H., 2008. Photographic guide to sample sites and ant
communities at Fort Benning, Georgia. http://facultyweb.berry.edu/jgraham/
Fort Benning Sites.html.
Graham, J.H., Freeman, D.C., Emlen, J.M., 1993. Developmental stability: a sensitive
indicator of populations under stress. In: Landis, W.G., Hughes, J.S., Lewis, M.A.
(Eds.), Environmental Toxicology and Risk Assessment. American Society for
Testing and Materials, Philadelphia, pp. 136–158.
Graham, J.H., Hughie, H.H., Jones, S., Wrinn, K., Krzysik, A.J., Duda, J.J., Freeman, D.C.,
Emlen, J.M., Zak, J.C., Kovacic, D.A., Chamberlin-Graham, C., Balbach, H., 2004.
Habitat disturbance and the diversity and abundance of ants (Formicidae) in
the Southeastern Fall-Line Sandhills. J. Insect Sci. 4, 30.
Graham, J.H., Krzysik, A.J., Kovacic, D.A., Duda, J.J., Freeman, D.C., Emlen, J.M., Zak,
J.C., Long, W.R., Wallace, M.P., Chamberlin-Graham, C., Nutter, J.P., Balbach, H.E.,
2008. Ant community composition across a gradient of disturbed military landscapes at Fort Benning, Georgia. Southeast Nat. 7, 429–448.
Graham, J.H., Raz, S., Hel-Or, H., Nevo, E., 2010. Fluctuating asymmetry: methods,
theory, and applications. Symmetry 2, 466–540.
Grime, J.P., 2001. Plant Strategies, Vegetation Processes, and Ecosystem Properties,
2nd ed. John Wiley & Sons Inc., New York, 456 pp.
Haywood, J.D., Tiarks, A.E., Sword, M.A., 1997. Fertilization, weed control, and pine
litter influence loblolly pine stem productivity and root development. New Forest 14, 233–249.
150
J.H. Graham et al. / Ecological Indicators 20 (2012) 143–150
Houle, G., Delwaide, A., 1991. Population structure and growth stress relationship
of Pinus taeda in rock outcrop habitats. J. Veg. Sci. 2, 47–58.
King, J.S., Albaugh, T.J., Allen, H.L., Kress, L.W., 1999. Stand-level allometry in Pinus
taeda as affected by irrigation and fertilization. Tree Physiol. 19, 769–778.
Kozlov, M.V., Niemelä, P., 1999. Difference in needle length—a new and objective
indicator of pollution impact on Scots pine (Pinus sylvestris). Water Air Soil Pollut.
116, 365–370.
Kozlov, M.V., Niemelä, P., Junttila, J., 2002. Needle fluctuating asymmetry is a sensitive indicator of pollution impact on Scots pine (Pinus sylvestris). Ecol. Ind. 1,
271–277.
Krzysik, A.J., Balbach, H.E., Duda, J.J., Emlen, J.M., Freeman, D.C., Graham, J.H., Kovacic,
D.A., Smith, L.M., Zak, J.C., 2005. Development of ecological indicator guilds for
land management. SERDP, RC-1114B, 119 pp.
Lozar, R.C., 2004. SEMP Historical Meteorology Evaluation for the Area Near Fort
Benning, GA: 1999–2001. U.S. Army Corps of Engineers, Engineer Research and
Development Center, ERDC/CERL Technical Note 04/1, 18 pp.
Maloney, K.O., Garten Jr., C.T., Ashwood, T.L., 2008. Changes in soil properties following 55 years of secondary forest succession at Fort Benning, Georgia, USA.
Restor. Ecol. 16, 503–510.
Marino, L.M., Oswald, B.P., Farrish, K.W., Williams, H.M., Unger, D.R., 2002. Growth
response of loblolly pine to intermediate treatment of fire, herbicide, and
fertilizer: preliminary results. U.S. Department of Agriculture, Forest Service,
Southern Research Station General Technical Report SRS-48, 31–34 pp.
Martin, T.A., Jokela, E.J., 2004. Stand development and production dynamics of
loblolly pine under a range of cultural treatments in north-central Florida, USA.
Forest Ecol. Manag. 192, 39–58.
Mason, J.M., 2003. Soil Survey of Russell County, Alabama. U.S. Department of Agriculture, Natural Resources Conservation Service, 202 pp.
McDonald, R.I., Urban, D.L., 2004. Forest edges and tree growth rates in the North
Carolina Piedmont. Ecology 85, 2258–2266.
NRCS, 2007. Web Soil Survey Version 1.1. Soil Survey Division, U.S. Department of
Agriculture. http://websoilsurvey.nrcs.usda.gov/app/WebSoilSurvey.aspx.
Oberhuber, W., Stumböck, M., Kofler, W., 1998. Climate-tree-growth relationships
of Scots pine stands (Pinus sylvestris L.) exposed to soil dryness. Trees-Struct.
Funct. 13, 19–27.
Ogle, K., Whitham, T.G., Cobb, N.S., 2000. Tree-ring variation in pinyon
predicts likelihood of death following severe drought. Ecology 81,
3237–3243.
Ozernyuk, N.D., Dyomin, V.I., Prokofyev, E.A., Androsova, I.M., 1992. Energy homeostasis and developmental stability. Acta Zool. Fenn. 191, 167–175.
Parsons, P.A., 2005. Environments and evolution: interactions between stress,
resource inadequacy and energetic efficiency. Biol. Rev. 80, 589–610.
Raz, S., Graham, J.H., Hel-Or, H., Pavlíček, T., Nevo, E., 2011. Developmental instability
of vascular plants in contrasting microclimates at ‘Evolution Canyon’. Biol. J. Linn.
Soc. 102, 786–797.
Rosenheim, J.A., Tabashnik, B.E., 1991. Influence of generation time on the rate of
response to selection. Am. Nat. 137, 527–541.
Smith, S.A., Donoghue, M.J., 2008. Rates of molecular evolution are linked to life
history in flowering plants. Science 322, 86–89.
Sword, M.A., Chambers, J.L., Gravatt, D.A., Haywood, J.D., Barnett, J.P., 1998. Ecophysiological response of managed loblolly pine to changes in stand environment.
In: Mickler, R.A., Fox, S. (Eds.), The Productivity and Sustainability of Southern
Forest Ecosystems in a Changing Environment. Springer-Verlag, New York, pp.
185–206.
Tamhane, A.C., 1979. A comparison of procedures for multiple comparisons of means
with unequal variances. J. Am. Stat. Assoc. 74, 471–480.
Tang, Z., Chambers, J.L., Guddanti, S., Barnett, J.P., 1999a. Thinning, fertilization, and
crown position interact to control physiological responses of loblolly pine. Tree
Physiol. 19, 87–94.
Tang, Z., Chambers, J.L., Guddanti, S., Yu, S., Barnett, J.P., 1999b. Seasonal shoot and
needle growth of loblolly pine responds to thinning, fertilization, and crown
position. Forest Ecol. Manag. 120, 117–130.
Zahner, R., Saucier, J.R., 1989. Twenty five years of climate related growth trends for
natural stands of loblolly pine in central Georgia, U.S. Department of Agriculture,
Forest Service, General Technical Report SO74, 25–32 pp.
Zutter, B.R., Miller, J.H., Allen, H.L., Zedaker, S.M., Edwards, M.B., Newbold,
R.A., 1998. Response of foliage of young loblolly pine to woody and
herbaceous plant control. In: Wagner, R.G., Thompson, D.G. (Eds.), Third International Conference on Forest Vegetation Management: Popular Summaries,
pp. 381–383.
Download